May 26, 1898] 



NATURE 



79 



in two quarters of the globe so widely separated as 

 Toronto and Hobarton presumably indicating not simply 

 a local effect, but one rather of cosmical character. He 

 pointed out that as the sun must be recognised as at 

 least the primary cause of all magnetic variations that 

 conform to a law of local hours, as does the solar diurnal 

 range, it seems not unreasonable to suppose that in the 

 case of other magnetic variations we should look, in the 

 first instance, to any periodical variation by which the sun 

 is aflFected, to ascertain whether any coincidence of period 

 or epoch is traceable. And he draws attention to the 

 circumstance that, according to Schwabe's then recently- 

 published table of frequency of solar spots, a minimum in 

 number of spots occurred ^in 1843 and a maximum in 

 1848, with progressive increase in the intermediate years 

 similar to that of the diurnal magnetic range during the 

 same interval as shown by the Toronto and Hobarton 

 observations. This led Sabine to infer the probable 

 existence of a periodical variation in magnetism similar 

 to that — one of about ten years — which Schwabe had 

 detected in sun-spots from observations extending over a 

 period of twenty-five years. 



In the meantime another worker had been busy with 

 the same subject. In Poggendorff's Annalen for Decem- 

 ber 185 1 there appeared the well-known paper by Dr. 

 Lamont, "on the ten-yearly period," in which he gave the 

 following values of diurnal range of the declination 

 magnet as observed at Munich. 



Lamont considered that these numbers indicated a 

 periodical variation, and from them he found by graphical 

 construction that a minimum apparently occurred in 

 1843 and a maximum in 1848. He further discussed 

 such older magnetic observations as were found to be 

 available, and came to the following conclusion, which it 

 may be interesting to give in his own words. " Die 

 grosse der Declinations-Variationen hat eine zehnjahrige 

 Periode, so zwar, dass sie mit regelmassigen Uebergange 

 fiinf Jahre im Zunehmen, und fiinf Jahre im Abnehmen 

 begrififen ist." 



Sabine became acquainted with Lamont's paper whilst 

 writing his own, and quotes Lamont's figures from 1843 

 to 1848, showing how the Munich results confirmed those 

 of Toronto and Hobarton. It would seem that Lamont and 

 Sabine each independently suspected the existence of a 

 periodical variation in diurnal magnetic range, which 

 Lamont appears to have first distinctly formulated in the 

 words quoted ; whilst it was to Sabine that the suggestion 

 that the periodical variation was one apparently concurrent 

 with that of sun-spots was due. Lamont considered the 

 variation to be so real that in any theory of the diurnal 

 movement it could not be disregarded. Sabine more 

 cautiously wrote : "As the physical agency by which the 

 phenomena are produced is in both cases unknown to us, 

 our only resource for distinguishing between accidental 

 coincidence and causal connection seems to '^^persever- 

 ance in observation, until either the inferences from a 

 possibly too limited induction are disproved, or until a 

 more extensive induction has sufficed to establish the 

 existence of a connection, although its precise nature may 

 still be imperfectly understood." In a postscript to 

 Sabine's paper (dated May 24, 1852) he gives a table of 

 mean diurnal range of declination for Toronto and 

 Hobarton from 1841 to 185 1, which clearly shows, as do 

 the Munich numbers, the minimum of 1843 ^"d the 

 maximum of 1848 ; and in 1856 he showed that at 

 Toronto, from 1844 to 1848, there was a progressive in- 

 crease in the amount of magnetic disturbance in all three 

 elements of declination and horizontal and vertical force. 



NO. 1 49 1, VOL. 58] 



Considering that the periodical variation of diurnal 

 range was found to exist in regions of the earth so far 

 apart as Toronto, Hobarton and Munich, the results at 

 the three places being distinctly corroborative, and,, 

 further, the circumstance that it appeared to be closely in 

 accord with the established solar-spot variation, it seems- 

 to be matter for reflection as to how it happened that irk 

 some quarters the agreement between the magnetic and 

 solar variations was thought to be only of apparent or 

 accidental nature. Sir George Airy, in his paper ' " On 

 the Diurnal Inequalities of Terrestrial Magnetism," had 

 occasion to give therein a list of the days of greater 

 magnetic disturbance at Greenwich in the years 1841 to 

 1857, and he incidentally remarks that "there is no 

 appearance of decennial cycle in their recurrence." But 

 this is not surprising, for although magnetic disturbance 

 does cluster about the epochs of maximum of sun-spots, 

 it is on occasions by no means closely confined thereto, 

 though nearly or quite absent at epochs of minimum of 

 sun-spots. Thus the periodical variation, as regards the 

 disturbance element, although existing, is not so distinctly 

 traceable unless longer periods are examined, accom- 

 panying sun-spot maxima as disturbance does in a 

 somewhat loose fashion as compared with the more 

 regular increase and decrease of diurnal magnetic range 

 with variation of sun-spot frequency. The behaviour of 

 magnetic disturbance in this respect is indeed a matter 

 that I am yet hoping to investigate more exactly. 



Then, again, Lamont appears to have adopted for the 

 diurnal magnetic range the difference between the 

 positions of the magnet at 8h. in the morning and ih. in 

 the afternoon, as being the times of the greatest easterly 

 and westerly deviation respectively. It is true that the 

 positions of the magnet at these hours would not be 

 likely to represent the extreme positions at Munich 

 throughout the year, especially as regards the easterly 

 deviation ; still the diurnal range resulting from the 

 employment of such fixed hours approximates in such 

 degree to the true range for Munich, as very well serves 

 clearly to bring out the decennial variation, of which 

 indeed the good agreement between Lamont's and 

 Sabine's results is of itself further proof, since the latter 

 do depend on observations extending through the twenty- 

 four hours of the day. From whatever cause, however, there 

 were those in earlier days who doubted the existence of 

 any real relation between magnetic and solar variations. 

 The so-called decennial period, it may be here mentioned, 

 seems to be more nearly an eleven-year period, this 

 being about the mean value, although it is variable in 

 length to the extent of several years. 



When, in the year 1875, ^ ^^s transferred at the Royal 

 Observatory from the Astronomical to the Magnetical 

 and Meteorological Department, I had then paid no- 

 particular attention to this question, and had an open 

 mind thereon. But the daily examination of the photo- 

 graphic records after a time convinced me that change 

 was in progress in the character of the records from year 

 to year, such as even in this simple daily inspection of the 

 records could not be well overlooked ; and acting involun- 

 tarilyon Sabine's principle of perseverance in observation, 1 

 came to the conclusion that it would be well to endeavour 

 further to investigate the facts of observation, especially 

 as the long series of Greenwich observations, made 

 throughout on the same general plan and with instru- 

 ments of the same kind, furnished so excellent an 

 opportunity for applying an independent test of the 

 reality or otherwise of the relation supposed to exist, 

 which the late Dr. Wolf, of Zurich, had already done so 

 much to establish. My first paper appeared in the 

 Philosophical Transactions for the year 1880, and 

 deals with the Greenwich observations from 1841 to 1877. 

 This I have recently supplemented by a second paper, 

 read before the Royal Society on March 10 of the 



1 Read before the Royal Society, April 23, 1863. 



